JP2000076110A - Recovering process system of decentralized file system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the system automatically recover in a short time by accessing a decentralized file by using location information on the file and restoring only a file judged to be broken from backup information. SOLUTION: When a user recognizes file breakage, a fault detecting and restoring process execution part 301 reads in the management information of the file by using a file operation part 304 and compares the identification record that the management information has with the value of an identification record storage part 303. When the identification record does not match the value or when the file can not be read in, it is judged that the file is broken. In this case, a backup information operation part 305 fetches the corresponding file from the backup information. Then the fault detecting and restoring process execution part 301 transfers the fetched file to the file operation part 304 to restore it in a directory on a specified host. Consequently, an efficient system recovery becomes possible.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a recovery processing system for a system for distributing and managing files on a computer system.

[0002]

2. Description of the Related Art When a failure occurs in a system for managing files in a distributed manner, it is normal to return all files to a state where backup has been executed in the past. This method has the disadvantage that the files that have survived the failure will be returned to the state where the backup was performed, and that the files that have been updated or added after the backup will be erased as a result of the disk initialization. There are often.

[0003] On the other hand, after a failure occurs, a method is considered in which a user manually recovers only a damaged file. However, in order to detect which file is damaged,
The user has to carry out various investigations, and has drawbacks such as taking a long time in proportion to the number of files and making a mistake.

[0004]

In the prior art, even if only some files in the system are damaged due to a failure,
In order to unconditionally restore all files to the state in which the backup was executed, recovery processing is also performed on files that have not been affected. On the other hand, when the user manually performs the recovery process, there is a problem that the recovery operation takes time in proportion to the amount of files to be managed. Further, there is a disadvantage that operations such as addition, update, and deletion performed after the backup are invalidated.

SUMMARY OF THE INVENTION An object of the present invention is to detect a file damaged due to a failure in response to a recovery request from a user after a failure occurs in the distributed file system, recover only that file from backup information, and quickly An object of the present invention is to provide a recovery processing system that realizes automatic recovery of a system.

[0006]

SUMMARY OF THE INVENTION The object of the present invention is to provide a means for storing file location information managed by a distributed file system comprising at least one computer, a plurality of file systems, and file backup information; A device is provided which has means for giving management information such as a record for determining that the file is managed and the file update time to each file, means for reading the management information, and means for recovering a specified file from backup information. Achieved by:

That is, according to the present invention, the distributed file is accessed using the location information of the file, and the file is managed by the system using the means for reading the management information given to the file. To determine if the file is corrupt. By using the means for recovering the specified file from the backup information and recovering only the file determined to be corrupt from the backup information, the system can be efficiently recovered.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In the present system, a specific record such as a product number and the time when a file is added or updated are added as management information to a file managed by the system at the time of addition or update. It has a mechanism to remove the management information and provide it.

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows the structure of a file managed by the present system. The file to be managed includes management information 101 and data 102. The management information 101 has an identification record 103 and an update time 104. In the identification record 103, a specific record such as a product number "H001" for determining that the file is managed by the system is written, and in the update time 104, the time at which the file was added or updated is recorded. The data 102 is the contents of the original file used by the user.

FIG. 2 shows location information in the present system. The location information is used to manage the location of which computer (host) stores the file when a specific file is specified when a plurality of computers are connected via a network. Information. The management information 101 described in FIG. 1, a file name 201 managed by the system, a host name 202 where the file exists, and a directory name 203
And a backup time 204 for recording the time at which the file was backed up.

FIG. 3 shows a file failure detection / recovery device, which is a system representing an embodiment of the present invention.
The failure detection / recovery device includes a failure detection / recovery processing execution unit 301 that manages operations associated with the failure detection processing and the recovery processing.
A location information operation unit 302 for operating location information, an identification record storage unit 303 for storing an identification record of management information, a file operation unit 304 for operating a file, and an operation for backup information. It comprises a backup information operation unit 305.

FIG. 4 shows an overall configuration for explaining an embodiment of the present invention.

The present system includes a file 401 managed by the system and its location information 402, a failure detection / recovery device 403 for detecting a corrupted file and recovering only the corrupted file from backup information, and a magnetic tape or the like. File backup information 404 using a medium, a plurality of file systems 406 for managing the file 401,
The computer 407 manages the location information 402, the failure detection / recovery device 403, and the backup information 404, and a network 405 connecting the plurality of file systems 406 and the computer 407.

FIG. 5 shows a file failure detection process. This process is started when the user recognizes that the file has been destroyed. Failure detection
The recovery processing execution unit 301 reads the file management information using the file operation unit 304, and compares the identification record of the read management information with the value of the identification record storage unit 303. If the identification records do not match or the file cannot be read, it is determined that the file is damaged. If the identification records match, the backup time of the location information is compared with the update time of the management information using the location information operation unit 302.

If the update time is later than the backup time, no further processing is performed and it is determined that the file is not damaged. If the backup time is later than the update time, it is considered that the read file and the file in the backup information are the same unless the file is destroyed. Therefore, the contents of the file read using the file operation unit 304 and the backup information operation unit 305 are used. Then, the contents of the file in the backup information 404 read using are compared to determine whether or not the file is damaged.

The file recovery processing, the file location information deletion processing, and the file location information recovery processing will be described below in order with reference to FIGS.

FIG. 6 shows a file recovery process. In the file recovery processing, the file determined to be damaged by the failure detection processing of FIG. 5 is recovered from the backup information 404. The flow of the file recovery process in FIG. 6 will be described with reference to the flowchart in FIG. Failure detection
The recovery processing execution unit 301 includes the backup information operation unit 30
5 is requested to take out the file determined to be broken (701). Backup information operation unit 3 that received the request
05 executes the extraction of the file from the backup information 404 (702). When the extraction processing is successful, the failure detection / recovery processing execution unit 301 transfers the extracted file to the file operation unit 304 (704).

The file operation unit 304 recovers the received file on the directory of the designated host (705). If there is no file determined to be corrupted in the backup information 404, the location information of the file is deleted (706).

FIG. 8 shows location information deletion processing (706) when there is no file determined to be corrupt in the backup information 404. The flow of the location information deletion processing of FIG. 8 will be described with reference to the flowchart of FIG. The failure detection / recovery processing execution unit 301 requests the location information operation unit 302 to delete the location information of the unrecoverable file (901). Upon receiving the request, the location information operation unit 302 deletes the information of the file from the location information 402 using the file name, the host name, and the directory name (902). This is a process for matching the fact that, in addition to the original file being destroyed, the backup file does not exist and only the location information remains, the present status is maintained.

FIG. 10 shows the location information recovery process. If it is determined that the location information 402 is broken, the management information of all the files is read and the location information 40 is read.
Recover 2. FIG. 1 shows the flow of the location information recovery process of FIG.
1 will be described. The failure detection / recovery processing execution unit 301 requests the file operation unit 304 to read management information of all files (1101). The file operation unit 304 reads the management information of each file (1102), and compares the identification record of the read management information with the value of the identification record storage unit (110).
3).

When the identification records match, the failure detection / recovery processing execution unit 301 determines that the file is managed by the system, and recovers the location information of the file as follows. The failure detection / recovery processing execution unit 301
The file operation unit 304 requests the name of the file, the name of the existing host, and the name of the directory as the location information (1104). The file operation unit 304 extracts the file name, the host name, and the directory name by using the function of the normal file system, and transmits the result to the failure detection / recovery processing execution unit 301 (110
5).

The failure detection / recovery processing execution unit 301 adds the result to the location information 402 using the location information operation unit 302 (1106). By executing the process for all the files, it is possible to recover the location information according to the state of the distributed file system after the occurrence of the failure. Further, the recovered location information 402 is compared with the location information in the backup information 404 (1107), and as a difference, a group of files that exist in the backup information 404 but do not exist in the recovered location information 402 is determined using a file recovery process. Can be recovered from the backup information 404 (11
08).

The flow of the failure detection / recovery processing in this system will be described with reference to the flowchart of FIG. After the occurrence of the failure, upon receiving a recovery request from the user, the failure detection / recovery device performs the failure detection processing described with reference to FIG. 5 on the location information (1201), and detects whether the location information has been destroyed. As a result, if the location information is not broken, a file recovery process is performed for each file managed by the location information (1202). In the file recovery processing, recovery from the backup information described in FIG. 6 or deletion of the location information described in FIG. 7 is executed.

If the location information has been destroyed, the location information recovery process described with reference to FIG. 8 is performed (1203), and the location information of the distributed file system is recovered from the files determined to be not corrupted. The recovered location information is compared with the location information in the backup information, and a file recovery process is performed based on the comparison result.

By using the processing method of the present invention, it is possible to detect a failed file and recover only the failed file in the distributed file system.

As a result, the recovery time of the system can be shortened as compared with the conventional processing method, and an inexpensive recovery processing system can be created as compared with the method of mirroring the entire system.

[0028]

According to the present invention, when a failure occurs in a system for distributing and managing files, the failed file is detected, the file is automatically recovered from the backup information, and the unbroken file is restored. Since the recovery process is not performed on the system, a more efficient system recovery can be realized as compared with a conventional recovery method such as a batch recovery method or manual recovery.

[Brief description of the drawings]

FIG. 1 is a diagram showing a distributed file structure and management information managed by the present system.

FIG. 2 is a diagram showing location information in the present system.

FIG. 3 is a system diagram showing a failure detection / recovery device according to an embodiment of the present invention.

FIG. 4 is a diagram showing an entire configuration of a management file system according to the embodiment of the present invention.

FIG. 5 is a diagram showing a file failure detection process.

FIG. 6 is a diagram showing a file recovery process according to the present invention.

FIG. 7 is a flowchart of a file recovery process according to the present invention.

FIG. 8 is a diagram showing location information deletion processing according to the present invention.

FIG. 9 is a flowchart of location information deletion processing according to the present invention.

FIG. 10 is a diagram showing location information recovery processing according to the present invention.

FIG. 11 is a flowchart of location information recovery processing according to the present invention.

FIG. 12 is a flowchart of a recovery process according to the present invention.

[Explanation of symbols]

101 management information, 102 data part, 10
3: identification record, 104: update time, 201: file name, 202: host name, 203: directory name, 204: backup time, 30
1: Failure detection / recovery processing execution unit 302: Location information operation unit 303: Identification record storage unit 304:
File operation unit, 305 ... backup information operation unit,
401 ... file, 402 ... location information,
403: Failure detection / recovery device, 404: Backup information, 405: Network, 406:
File system, 407 ... computer.

 ──────────────────────────────────────────────────の Continuing on the front page (72) Inventor Kimitaka Tamura 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Inside Software Division, Hitachi, Ltd. (72) Inventor Yoichi Haga 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa F-term in Hitachi, Ltd. Software Division (Reference) 5B045 DD16 JJ43 5B065 EA15 EA24 5B082 DC02 DC07 DD06 DE01 DE06

Claims (4)

[Claims] In a distributed file system comprising at least one computer, a plurality of file systems, and file backup information, each file has location information of files managed by the system and management information for each file. A recovery system in a distributed file system, comprising means for detecting a failed file using file management information and recovering the failed file from backup information. 2. A system according to claim 1, wherein the management information such as a record for determining that the file is managed by the system and the update time of the file is given to the file managed by the system, and the management information is read. A recovery processing system in a distributed file system, wherein a failed file is detected. 3. The method according to claim 1, wherein when the location information of the file managed by the system is destroyed, the presence or absence of the destruction is determined by a method, the host name and directory name of the unbroken file are extracted, and the contents are extracted. A recovery processing system in a distributed file system, wherein location information is recovered by using the same. A means for storing location information of a file managed by the system in the computer; a means for providing management information to each file; a means for detecting a failed file using the management information of each file; A recovery processing system in a distributed file system, which is a machine-readable storage medium storing a program for executing means for recovering a received file from backup information.